This proposal builds on a rich resource of bladder cancer (BC) cases and controls derived from two ongoing BC studies at the University of Texas M.D. Anderson Cancer Center, and from two large independent U.S. BC studies - the New England BC Study and the New Hampshire BC Study. The goal is to identify genetic loci that predispose individuals to BC through a genome-wide scanning approach. There are five specific aims. Aim 1 is to first perform genome-wide, high-density SNP genotyping using the Illumina HumanHap550 SNP platform on 800 cases and 800 controls from M.D. Anderson, with a target of candidate SNPs of about 28,000;followed by an internal validation to narrow down candidate SNPs to about 1,536 using additional 800 pairs of cases and controls. Illumina's Custom Infinium array will be the genotyping format. In this aim, in addition to individual SNP analysis, we will also implement haplotype-based analyses and pathway aggregation analysis to identify additional genetic loci that might have been overlooked using individual SNP analysis. Aim 2 is the first external validation of the 1,536 SNPs from Aim 1 using 1,000 pairs of cases and controls from the New England BC Study. Illumina's GoldenGate assay will be the genotyping format. After this stage, the candidate SNPs will be narrowed down to about 100. Aim 3 is the second independent external validation using 750 pairs of cases and controls form the New Hampshire BC Study. In this aim, the 100 top candidate SNPs passed from Aims 1 and 2 plus additional functional SNPs in genes containing these SNPs will be genotyped. GoldenGate assay will be used for this aim. Aim 4 is to perform fine mapping studies in the flanking regions of each of the top 25 SNP loci confirmed in Aim 3 to identify causative loci. This will utilize all 6,700 cases and controls. An average of 15 additional SNPs (tagging SNPs and functional SNPs) per gene is expected. Aim 5 is to apply novel machine-learning tools to identify any gene-environment and gene-gene interactions greatly influencing risk for BC in all the 6700 subjects. These analyses will be utilized to examine SNP main effect and develop and validate algorithms that will identify individuals at highest risk for BC, given their personal exposure patterns and their genetic risk profiles. This proposal applies state of art technology to perform a multistage, genome-wide SNP analysis in three largest, well-characterized U.S. population of BC cases and controls, and incorporates complete epidemiologic data and rich and unique functional data. In addition, results from this study will be provided to the International Consortium of BC Case Control Studies for future validation. The ability to identify genetically susceptible, high-risk subgroups that would benefit from intensive screening and/or chemopreventive interventions is of immense clinical and public health benefit. PUBLIC HEALTH RELEVANCE Bladder cancer (BC) is a disease mainly caused by smoking and occupational exposure. However, only a small percentage of exposed individuals develop BC. Inherited host genetic factors may play an important role in determining an individual's susceptibility to BC. This proposal builds on three largest well-characterized U.S. populations of BC cases and controls - the M. D. Anderson BC study, the New England BC Study and the New Hampshire BC Study. A total of 6700 cases and controls will be used. The goal is to identify genetic loci that predispose individuals to BC through a non-biased, discovery- driven, genome-wide scanning approach and to incorporate complete epidemiologic data and rich and unique functional data. About 550,000 genetic variations of human genome will be initially screened in an M.D. Anderson population consisting of an equal number of BC cases and normal controls. About 28,000 top candidate variations that are potentially associated with increased BC risk will be first narrowed down and internally replicated in a second M.D. Anderson population, and then be validated in the other two independent U.S. BC populations. Finally, the causative genetic loci that predispose individuals to BC will be mapped. This study is significant because by identifying BC susceptibility loci, it will shed light into the biological mechanisms of BC etiology. Furthermore, it may facilitate identifying high-risk subgroups of individuals for BC, given their genetic makeup and environmental exposures. The ability to identify high-risk subgroups of individuals for BC will provide immense public health benefit for those high- risk people who may be subjected to close surveillance and chemoprevention.